customizing cancer immunotherapies to match the intrinsic

Customizing cancer immunotherapies

to match the intrinsic tumor

microenvironment

Brad Nelson, PhD

British Columbia Cancer Agency

Precision Medicine Retreat,

August 9 2017

The cancer genome is an altered version of self

Spectral karyotype analysis

Normal Cancer

1 2 3 4 5

6 7 8 9 10 11

12 13 14 15 16 17 18

19 20 21 22 X

1 2 3 4 5

6 7 8 9 10 11

12 13 14 15 16 17 18

19 20 21 22 X

1 2 3 4 5

6 7 8 9 10 11

12 13 14 15 16 17 18

19 20 21 22 X

1 2 3 4 5

6 7 8 9 10 11

12 13 14 15 16 17 18

19 20 21 22 X

1 2 3 4 5

6 7 8 9 10 11 12

13 14 15 16 17 18

19 20 21 22 X

1 2 3 4 5

6 7 8 9 10 11

12 13 14 15 16 17 18

19 20 21 22 X

A B

C D

E F

No two cancers are alike

Chromosomes from 6 different breast cancers

Immune recognition of cancer:

Tumor-infiltrating lymphocytes (TIL)

CD8+ killer T cells

CD20+ B cells

CD4+ T cells

Tumor cells

Multi-colour IHC of high-grade serous ovarian cancer (HGSC)

Katy Milne

TIL are strongly associated with survival

in human cancer

100

50

0 % o

ve

rall s

urv

ival

Years

Dense CD8+ TIL Sparse CD8+ TIL

BCCA/VGH cohort

high-grade serous (HGSC)

optimally de-bulked

n = 200

p = 0.0008

5 10 15 0 Clarke, B. et al. 2009

Milne, K. et al. 2009

HGSC shows extensive

intratumoral heterogeneity

A McPherson...S Shah, Nature Genetics 2016




How does the immune

system contend with ITH?




How does the immune


• the players

• the strategy

The strongest TIL responses

involve both T cells and B cells

Few TIL

Three cases of HGSC:

Weak TIL

T cells Robust TIL

T cells and B cells

CD4+ T cells

CD8+ T cells

CD20+ B cells

T cells and B cells show a

combined effect on survival

Kaplan-Meier based on TIL patterns in HGSC (n=167, optimally de-bulked)

Julie Nielsen et al, Clin Can Res 2012

CD3+ T cells

CD20+ B cells

CD208+ dendritic cells

PNAd+ high

endothelial venules

CD8+ T cells

CD21+ follicular DCs

David Kroeger et al, Clin Can Res 2016

Strong TIL responses involve

Tertiary Lymphoid Structures (TLS)

DC

Y

Direct effects Complement ADCC

Plasma cell

Cytokines Chemokines

Enhanced antigen presentation

Optimal TIL responses involve both

cytolytic and antibody-based mechanisms

B cell boost

Tumor cell CD8+ killer T cell

CD4+ helper T cell

David Kroeger et al, Clin Can Res 2016




How does the immune


• the players

• the strategy

Do TIL recognize truncal or branch

features of the tumor phylogeny?

1. The players

2. The strategy

Do TIL recognize truncal or branch

features of the tumor phylogeny?

Extensive spatial profiling of

120 tumors from 21 HGSC patients

Allen Zhang, Sohrab Shah, in preparation

Allen Zhang, Sohrab Shah, in preparation

T-cell clones appear to track

with tumor clones across space

Tumor clonal distance

Patient 1 Patient 2 Patient 3

Patient 4 Patient 9 Patient 10

TIL may recognize branch features

of the tumor phylogeny

Different tumor types harbour different

immune “communities” (and strategies?)

• RNA-seq data from 21 cancer

types in TCGA (7,893 cases)

• xCell used to estimate abundance

of 38 immune cell types

• TIL “communities” are projected in

2D using tSNE

Phineas Hamilton, unpublished

Few TIL


Weak TIL

T cells Robust TIL

T cells and B cells

CD4+ T cells

CD8+ T cells

CD20+ B cells

How can we best enhance

anti-tumor immunity?

The PD-1 pathway has emerged as

a major control point in anti-tumor

immunity

The PD-1 pathway has emerged as

a major control point in anti-tumor immunity

PD-L1 is expressed in TIL-rich tumor regions

CD8+ T cells

PDL1+ cells

John Webb et al, Can Imm Res 2015, Gyn Onc, 2016

Checkpoint blockade releases

the brakes on anti-tumor immunity

anti-CTLA-4 (eg, Ipilimumab)

• Metastatic melanoma – FDA approval

anti-PD-1 (eg, Nivolumab, Pembrolizumab, others):

• Metastatic melanoma – 38% Objective Responses (Hamid, NEJM 2013), 53%

Objective Responses with Ipilimumab (Wolchok, NEJM 2013) and FDA approval

• Non-small cell lung cancer – 18% Objective Responses and FDA approval

• Kidney cancer – 27% Objective Responses (Topalian, NEJM 2012); 52% ORR

nivolumab + sunitinib (Amin, JCO abstract, 2014), FDA approval

• Bladder cancer – 52% Objective Responses (Powles, Nature 2014), FDA approval

• Hodgkin’s Lymphoma – 87% Objective Responses (Ansell, NEJM 2015), FDA

approval

• Colorectal cancer (MSI) – 40% Objective Responses (Le, NEJM 2015), FDA

Breakthrough Status 2015

• Any adult or pediatric metastatic solid tumor with mismatch repair deficiency (dMMR),

FDA approval

• Replacing frontline chemotherapy for melanoma and lung cancer (so far)

Checkpoint blockade: clinical successes

Cost

• approx. $100k/treatment cycle

• combinations may be required for some cancers (e.g.

Ipi + Nivo for melanoma)

• long-term use may be required for some cancers

Efficacy

• many cancers (e.g. ovarian, breast) have low

response rates (10-20% range)

• responses are often transient (e.g. lung)

Checkpoint blockade: current challenges

Stimulatory and inhibitory pathways in T cells

T cell Antigen presenting cell

or tumor cell

Few TIL


Weak TIL

T cells Robust TIL

T cells and B cells

CD4+ T cells

CD8+ T cells

CD20+ B cells

How can we best enhance

anti-tumor immunity?

Cold tumors exhibit profound

lymphocyte infiltration barriers

CD8+ killer T cells

CD20+ B cells

CD4+ T cells

Tumor or

blood sample

Identify/engineer

tumor-reactive T cells

Expand T cells

Infuse T cells

with immune

modulation

Adoptive T cell therapy (ACT)

Clinical grade T cell production unit BCCA’s Deeley Research Centre, Victoria

Antigens

Access

Activity

Keys to successful T cell therapy

Targeting driver mutations in lymphoma

1. Collect tumour samples

BCCA and affiliated hospitals

2. Identify mutations

Michael Smith Genome Sciences Centre

3. Assess immunogenicity of mutations

BCCA’s Deeley Research Centre

4. Vaccinate patients and assess

clinical outcomes

BCCA Clinical Trials Unit

Patient’s tumor biopsy at relapse

Sequence panel of 50 genes to

identify driver mutations

Identify and expand mutation-specific

CD4 and/or CD8 T cells

Infusion with immune modulation

Team:

Julie Nielsen, PhD

Nicol MacPherson, MD

Laurie Sehn, MD

Joe Connors, MD

Raewyn Broady, MD

John Webb, PhD

Ryan Morin, PhD

Brad Nelson, PhD

MYD88

EZH2

MEF2B

CREBBP

EP300

CARD11

MLL2

PIM1

FOXO1

IRF4

CD79B

etc.

Julie Nielsen et al, Clin Can Res 2016, Oncoimmunology 2017

Chimeric Antigen Receptors (CARs)

Antibody portion (e.g. a-CD19)

Spacer and

transmembrane

domain

Co-stimulatory domain

from CD137 or CD28

T cell receptor

signaling domain

CD19 CAR-T cells:

• 90% Complete Responses (67% sustained) in pediatric and adult

leukemia (Davila, Sci Trans Med 2013; Maude, NEJM 2014)

• FDA Breakthrough Designation pediatric and adult leukemia in 2014

• 80% Objective Responses in lymphoma (Kochenderfer, JCO 2014)

Victoria

Vancouver

Canadian CAR-T Cell Network

• Vector development (Holt, Yung)

• T cell production (Webb, Nelson)

• Clinical trial site (Broady)

Ottawa

• Clinical-grade virus production (Bell)

• Socioeconomic (Ferguson)

• Clinical trial site (Atkins, Kekre)

* * * *

Spacer

Spacer

Targeting element

(e.g. single-chain variable fragments,

scFvs)

Transmembrane domain (e.g. CD8 alpha)

Costimulatory domain

(e.g. 4-1BB or CD28)

Activation domain

(e.g. CD3 zeta)

Non-covalent interactions *

Dimerization domain A

(e.g. Jun leucine zipper)

Dimerization domain B

(e.g. Fos leucine zipper)

With Marty Boulanger (UVic):

Mutual Antibody T Cell Engagers (MATEs)

Engineering logic gates to enable T cells

to recognize “constellations” of antigens

Adapted from Davies and Maher, Trans Can Res 2016

Antigens • driver mutations

• cell surface antigens (CARs, MATEs)

Access

Activity


Cold tumors exhibit profound

lymphocyte infiltration barriers

CD8+ killer T cells

CD20+ B cells

CD4+ T cells

Circumventing infiltration barriers

using Cbl-b deficient CD8+ T cells

Taimei Yang et al, CII 2009

Adoptive T cell therapy of mammary tumors

using CD8+ OT-I T cells

WT Cbl-b -/-

WT Cbl-b -/-

Cbl-b



Access • enhanced T cell receptor signaling (cbl-b -/-)

Activity


With Marty Boulanger (UVic), Surjit Dixit (Zymeworks):

Native G-CSF:G-CSFR signaling complex

Y

Y

STAT5

SHC

JAK3

JAK1

JAK2

JAK2

IL-2 signal G-CSF signal

G-CSFmut:G-CSFRmut

exclusive heterodimer

Clinician-controlled cytokine signaling

using exclusive cytokine:receptor pairs

Mutant cytokine:receptor complex



Access • enhanced T cell receptor signaling (cbl-b -/-)

Activity • clinician-controlled cytokine receptors


The cancer genome is an altered version of self

Normal Cancer

Re-wiring T cells to recognized altered self

Engineered T cell Cancer

Genomics - Vancouver:

Rob Holt, PhD

Scott Brown

Ryan Morin, PhD

Sohrab Shah, PhD

Allen Zhang

Gregg Morin PhD

Steven Jones PhD

Funding:

BC Cancer Foundation

CCSRI

CIHR

TFRI

US DOD

CRS

Genome BC

IWMF

LLSC

BioCanRx NCE

Conconi Family

Special thanks to our patients

Nelson Lab:

Maartje Wouters, PhD

Alex Rodriguez, PhD

Phineas Hamilton, PhD

Stephen Redpath, PhD

Nicole Little

Julian Smazynski

Luke Neufeld

Angela Cheng

Meghan Hand

Eunice Kwok

Megan Fuller

Collaborators - Vancouver:

Anna Tinker, MD

Raewyn Broady, MD

Blake Gilks, MD

David Huntsman, MD

Jessica McAlpine, MD

Dianne Miller, MD

Michael Anleso,

Randy Gascoyne, MD

Joe Connors, MD

Laurie Sehn, MD

Collaborators - Victoria:

Peter Watson, MD

Julian Lum, PhD

Nicol MacPherson, MD

Brian Berry, MD

Jodi LeBlanc, RN

Sindy Babinsky

Ellissa McMurtrie, MD

Mona Mazgani, MD

Immunotherapy Program:

Rob Holt, PhD

John Webb, PhD

Julian Lum PhD

Julie Nielsen, PhD

David Bond, PhD

Victoria Hodgson

Leah McCormick

Evelyn Wiebe

Mhairi Sigrist, PhD

Miruna Bala

Ana Subramanian

MCIC (Histo Core):

Katy Milne

Heather Derocher

Bronwyn Gibson-Wright

Sonya Laan

Stacey LeDoux

Key Alumni:

Darin Wick

David Kroeger, PhD

Ron deLeeuw, PhD

Charlotte Lo

Spencer Martin

Colin Sedgwick

Kwame Twumasi-

Boateng, PhD

customizing cancer immunotherapies to match the intrinsic

Documents